Impact of paravalvular regurgitation on the mid-term outcome after transcatheter and surgical aortic valve replacement

Abstract

OBJECTIVES

Open in new tabDownload slide

The aim of this study was to evaluate the incidence and prognostic impact of paravalvular regurgitation (PVR) on the outcome after transcatheter (TAVR) and surgical aortic valve replacement (SAVR) for aortic stenosis.

METHODS

The nationwide FinnValve registry included data on 6463 consecutive patients who underwent TAVR (n = 2130) or SAVR (n = 4333) with a bioprosthesis for the treatment of aortic stenosis during 2008–2017. The impact of PVR at discharge after TAVR and SAVR on 4-year mortality was herein investigated.

RESULTS

The rate of mild PVR was 21.7% after TAVR and 5.2% after SAVR. The rate of moderate-to-severe PVR was 3.7% after TAVR and 0.7% after SAVR. After TAVR, 4-year survival was 69.0% in patients with none-to-trace PVR, 54.2% with mild PVR [adjusted hazard ratio (HR) 1.64, 95% confidence interval (CI) 1.35–1.99] and 48.9% with moderate-to-severe PVR (adjusted HR 1.61, 95% CI 1.10–2.35). Freedom from PVR-related reinterventions was 100% for none-to-mild PVR and 95.2% for moderate-to-severe PVR. After SAVR, mild PVR (4-year survival 78.9%; adjusted HR 1.29, 95% CI 0.93–1.78) and moderate-to-severe PVR (4-year survival 67.8%; adjusted HR 1.36, 95% CI 0.72–2.58) were associated with worse 4-year survival compared to none-to-trace PVR (4-year survival 83.7%), but the difference did not reach statistical significance in multivariable analysis. Freedom from PVR-related reinterventions was 99.5% for none-to-trace PVR patients, 97.9% for mild PVR patients and 77.0% for moderate-to-severe PVR patients.

CONCLUSIONS

This multicentre study showed that both mild and moderate-to-severe PVR were independent predictors of worse survival after TAVR. Mild and moderate-to-severe PVR are not frequent after SAVR, but tend to decrease survival also in these patients.

Clinical trial registration number

ClinicalTrials.gov Identifier: NCT03385915.

INTRODUCTION

Transcatheter aortic valve replacement (TAVR) has emerged as the preferred method over surgical aortic valve replacement (SAVR) for the treatment of aortic stenosis (AS) in patients with high operative risk and is a treatment option also for lower risk patients [1, 2].

Paravalvular regurgitation (PVR), particularly when moderate or severe, is recognized as a significant complication occurring after TAVR [2–5]. Instead, the prognostic significance of mild PVR is less clear [3, 6, 7]. Previous studies reported rates of moderate-to-severe PVR ranging from 3.5% to 12%, while mild PVR rates are much higher and may range from 20% to 40%, depending on the type of prosthesis and the characteristics of the patient population [1, 8–14]. The occurrence of mild PVR remains high after TAVR in comparison to SAVR. In the intermediate-risk patients of the PARTNER 2 study, the rate of mild PVR at 30-day was 22.5% after TAVR and 2.8% after SAVR, whilst the rate of moderate-to-severe PVR was 3.7% after TAVR and 0.6% after SAVR [9]. Understanding the clinical significance of mild PVR is important because TAVR is increasingly performed in low-risk patients with long life expectancy. In this nationwide study, we investigated the prognostic impact of PVR on the mid-term outcome after TAVR and SAVR in an unselected patient population.

PATIENTS AND METHODS

Study data

The FinnValve registry (ClinicalTrials.gov Identifier: NCT03385915) is a nationwide, retrospective study including data from consecutive and unselected patients who underwent TAVR or SAVR with a bioprosthesis for severe AS at all 5 Finnish University Hospitals (Helsinki, Kuopio, Oulu, Tampere and Turku) from January 2008 to October 2017. The study protocol was approved by the Institutional Review Board of each participating centre.

The inclusion criteria for the study entry were as follows: (i) AS with or without aortic regurgitation; (ii) patients aged >18 years; and (iii) primary TAVR or SAVR with a bioprosthesis with or without concomitant coronary revascularization. The exclusion criteria were as follows: (i) any prior TAVR or surgical intervention on the aortic valve; (ii) concomitant major cardiac procedure on the ascending aorta and/or other heart valves or structures; (iii) patient operated on for isolated aortic regurgitation; (iv) heart valve endocarditis; or (v) SAVR with mechanical valve prosthesis. The operative risk of all patients was stratified according to the Society of Thoracic Surgeons (STS score) [15] and the European System for Cardiac Operative Risk Evaluation (EuroSCORE) II [16] risk scoring methods.

Data were collected retrospectively into an electronic case report form by cardiologists, cardiac surgeons and trained research nurses, and underwent robust checking of its completeness and quality. Data on mortality were retrieved from the national registry Statistics Finland, which collects data on death certificates reviewed by local and central authorities. Follow-up was considered complete for all patients except for those not residing in Finland, and their follow-up was truncated at hospital discharge.

Definition criteria of baseline risk factors

Baseline variables were defined according to the EuroSCORE II criteria [15]. Coronary artery disease was defined as any stenosis ≥50% of the main coronary branches. Anaemia was defined as baseline haemoglobin <120 g/l for women and <130 g/l in men. Chronic kidney disease classes 3–5 were defined as an estimated glomerular filtration rate <60 ml/min/1.73 m².

PVR was graded at hospital discharge at each centre by experienced cardiologists with transthoracic echocardiography into none-to-trace, mild, moderate and severe PVR (Supplementary Material, Table S1) [17, 18]. Patients with moderate and severe PVR were grouped together because of the small number of severe PVR (3 after TAVR and 9 after SAVR).

Outcome measures

The primary outcome was 4-year all-cause mortality. The secondary outcomes were stroke, blood transfusion, major vascular complications, reoperation for bleeding, infectious complications, acute kidney injury and atrial fibrillation. Late secondary outcomes included reinterventions on the implanted aortic bioprosthesis. Vascular complications were defined according to the Valve Academic Research Consortium-2 consensus document (VARC-2) [18]. Acute kidney injury was defined according to the KDIGO classification criteria (Supplementary Material, Table S2) [19]. Severe bleeding was defined as European Coronary Artery Bypass Grafting (E-CABG) bleeding grades 2–3, i.e. transfusion >4 units of red blood cells and/or reoperation for bleeding (Supplementary Material, Table S3) [20]. Infectious complications were deep sternal wound infection or mediastinitis, vascular access site infection, pneumonia, sepsis or other defined infection.

Statistical analysis

Categorical variables are presented as counts and percentages were compared using the χ² test. Continuous variables are presented as the mean and standard deviation or median and 25th–75th interquartile range, and were compared using the Student’s t-test or Mann–Whitney U-test based on their distributions. Comparisons involving patients with PVR of different severity were performed using the Jonckheere–Terpstra test. Time-trend analysis was done with linear-by-linear association test. Kaplan–Meier estimates were used to estimate survival, and comparisons were performed using the log-rank test.

A forward stepwise multivariable analysis with a binary logistic regression method was performed to identify independent risk factors for PVR. Regression models included covariates with P-value <0.1 in univariate analysis, i.e. age, body mass index, estimated glomerular filtration rate, maximal aortic gradient, gender, diabetes, bicuspid aortic valve, combined aortic valve disease, severe coronary artery disease, prior percutaneous coronary intervention, extracardiac arteriopathy, porcelain aorta, frailty, atrial fibrillation, systolic pulmonary artery pressure, left ventricular ejection fraction <51% and prior cardiac surgery. Operative variables of TAVR included valve types, balloon- or self-expandable valves and predilatation or postdilatation.

Independent predictors for late mortality were identified with the Cox proportional hazards method. The following variables with P-value <0.1 in univariate analysis were included in the regression model: PVR, age, body mass index, estimated glomerular filtration rate, anaemia WHO criteria, diabetes, pulmonary disease, atrial fibrillation, extracardiac arteriopathy, porcelain aorta, Geriatric Status Scale grades 2–3, systolic pulmonary artery pressure, prior cardiac surgery, combined aortic valve disease, bicuspid aortic valve, left ventricular ejection fraction <51%, severe coronary artery disease, major vascular complications, acute kidney injury, infectious complications and red blood cell transfusions or E-CABG bleeding severity grades 2–3. P-value <0.05 was set for statistical significance. Statistical analyses were performed using SPSS version 25.0 statistical software (IBM Corporation, New York, NY, USA).

RESULTS

The FinnValve registry includes data on 6463 patients who underwent primary TAVR or SAVR with bioprosthesis: 2130 (33.0%) patients underwent TAVR and 4333 (67.0%) underwent SAVR. Data on PVR were available in 2129 TAVR patients and 4330 SAVR patients, and these patients are the subjects of this study.

The mean follow-up time was 2.9 ± 1.8 years in the TAVR cohort and 5.0 ± 2.8 years in the SAVR cohort. Baseline characteristics of patients are shown in Table 1.

Transcatheter aortic valve replacement

Incidence of paravalvular regurgitation

The incidences of mild and moderate-to-severe PVR after TAVR were 21.7% and 3.7%, respectively. The rates of PVR declined significantly during the last study period (P < 0.001) (Fig. 1). The rates of PVR for different TAVR prostheses are reported in Fig. 2.

Risk factors for paravalvular regurgitation

The baseline characteristics of TAVR patients stratified by PVR severity are summarized in Supplementary Material, Table S4. In multivariable analysis, maximal transvalvular gradient, combined aortic valve disease, systolic pulmonary pressure >55 mmHg, self-expanding valve and predilatation of the native valve were independent risk factors for moderate-to-severe PVR after TAVR (Table 2). Newer generation transcatheter prostheses (Sapien 3, Evolut R and Pro, Lotus and Acurate Neo) were associated with significantly lower risk of PVR in comparison to early-generation prostheses (Sapien, Sapien XT, Corevalve) [odds ratio (OR) 0.23, 95% confidence interval (CI) 0.14–0.40] (Table 3). These newer devices were introduced on January 2014 and their use coincided with a decrease of PVR since then (Fig. 1). Age (OR 1.04, 95% CI 1.02–1.06) and prior cardiac surgery (OR 1.31, 95% CI 1.02–1.70) were additional risk factors for mild-to-severe PVR. It is worth noting that TAVR was performed in 114 patients with bicuspid aortic valve and it was not associated with a higher rate of mild-to-severe PVR (21.1% vs 25.7%, P = 0.27) or moderate-to-severe PVR (2.6% vs 3.8%, P = 0.80) compared with tricuspid aortic valves. When newer generation TAVR devices were evaluated, mild-to-severe PVR was observed in 2.2% of patients with tricuspid aortic valve patients and in none with bicuspid aortic valve (P = 0.25). Furthermore, patients who received a permanent pacemaker within 30 days after TAVR did not have an increased or decreased risk of mild-to-severe PVR (22.2% vs 25.8%, P = 0.71) or moderate-to-severe PVR (4.9% vs 3.6%, P = 0.39).

Survival and reinterventions

Early outcomes are summarized in Supplementary Material, Tables S5 and S6. After TAVR, 4-year survival of patients with none-to-trace PVR was 69.0%, with mild PVR 54.2% (P < 0.001) and with moderate-to-severe PVR 48.9% (P < 0.001). In multivariable analysis, mild PVR [adjusted hazard ratio (HR) 1.64, P < 0.001, 95% CI 1.35–1.99] and moderate-to-severe PVR (adjusted HR 1.61, P = 0.01, 95% CI 1.10–2.35) were independent risk factors for 4-year all-cause mortality (Table 3).

Freedom from all reinterventions on the aortic valve prosthesis was 99.6% in patients with none-to-trace PVR, 100% with mild PVR and 93.3% with moderate-to-severe PVR (P < 0.001). Freedom from PVR-related reinterventions was 100% for none-to-mild PVR and 95.2% for moderate-to-severe PVR (P < 0.001). Four patients underwent reintervention after TAVR because of moderate-to-severe PVR (2 percutaneous PVR closure procedures with a plug device, 1 valve-in-valve TAVR and 1 extensive surgical operation including SAVR, tricuspid valvuloplasty and closure of a perimembranous ventricular septal defect which caused significant left-to-right shunt and PVR). Patients with mild PVR did not undergo any repeat procedure on the aortic prosthesis. Three non-PVR-related reinterventions were performed in the none-to-trace PVR group.

Surgical aortic valve replacement

Incidence of paravalvular regurgitation

The rates of mild PVR and moderate-to-severe PVR after SAVR were 5.2% and 0.7%, respectively, and these did not change significantly during the study period (data not shown).

Risk factors for paravalvular regurgitation

In multivariable analysis, male gender, porcelain aorta and prior cardiac surgery were risk factors for moderate-to-severe PVR (Table 2). For mild-to-severe PVR, bicuspid aortic valve was a risk factor for PVR (P = 0.01; OR 1.42, 95% CI 1.07–1.89) along with male gender (P = 0.007; OR 1.44, 95 CI 1.11–1.87). PVR was not affected by different types of SAVR prostheses (data not shown). Patients who received a permanent pacemaker within 30 days after SAVR did not have an increased risk of mild-to-severe PVR (7.8% vs 5.9%, P = 0.88) or moderate-to-severe PVR (0.6% vs 0.7%, P = 1.0).

Survival and reinterventions

Early outcomes are summarized in Supplementary Material, Table S5. After SAVR, 4-year survival of patients with none-to-trace PVR was 83.8%, with mild PVR was 78.9% and moderate-to-severe PVR was 67.8% (P = 0.002) (Fig. 3). After adjusting for multiple covariates, mild and moderate-to-severe PVR were not associated with poorer 4-year survival (Table 4).

Freedom from all reinterventions on the aortic valve was 99.5% in patients with none-to-trace PVR, 97.9% with mild PVR (P = 0.01) and 77% with moderate-to-severe PVR (P < 0.001). Freedom from PVR-related reinterventions was 99.8% for none-to-trace PVR, 99.1% for mild PVR (P = 0.01) and 77% for moderate-to-severe PVR (P < 0.001). Sixty (1.5%) patients with none-to-trace PVR underwent late reintervention on the aortic valve prosthesis. Six of these reinterventions were performed because of worsening PVR. Eight (3.5%) patients with mild PVR underwent reintervention, 2 of them for PVR, 4 for endocarditis and 2 for structural valve degeneration. Eight (19.5%) patients with moderate-to-severe PVR underwent reintervention and all these procedures were related to PVR. These reinterventions included 1 TAVR, 5 SAVRs and 1 Bentall–DeBono procedure with CABG. One patient had a second reintervention for a new onset of PVR, which was treated percutaneously with a plug.

DISCUSSION

The present study demonstrated that both mild and moderate-to-severe PVR after TAVR were associated with impaired mid-term survival and that PVR rates diminished over time. Mild and moderate-to-severe PVR were uncommon after SAVR and the limited number of patients with this complication prevented conclusive results on the prognostic impact of PVR. However, the adjusted risk estimates suggest that mild and moderate-to-severe PVR tended to result in a decreased 4-year survival, also among SAVR patients.

Mild PVR after TAVR was associated with higher mortality during follow-up. Conflicting results on this issue have been reported by previous studies. In the PARTNER trial, the rate of mild PVR was 38%, and it was associated with increased 1-year mortality when compared to none-to-trace PVR (22.2% vs 15.9%) [6]. In the same trial, the rate of moderate-to-severe PVR was 9.1% with 1-year mortality of 35.1% [6]. A meta-analysis demonstrated higher all-cause mortality in patients with mild PVR compared to none-to-trivial PVR (HR 1.26, 95% CI 1.11–1.43; P < 0.001) [4]. Still, several studies did not confirm the negative prognostic impact of mild PVR after TAVR [5, 11, 21, 22].

We identified a number of risk factors associated with moderate-to-severe PVR after TAVR (Table 2). Most of these risk factors are likely an indirect measure of the severity of AS, i.e. transvalvular gradient and increased pulmonary artery pressure. Also a predilatation strategy was associated with increased risk of PVR, but this was most likely performed in patients with asymmetric or severe aortic valve calcification. Importantly, this study confirmed that the use of newer TAVR devices introduced in 2014 was associated with a significant decrease in PVR rates (Fig. 1).

Inconsistencies in categorizing PVR in different trials have been an obvious limitation in understanding the relationship between mild PVR and mortality after TAVR. A five-class scheme has been proposed by Pibarot et al. [23] to allow ‘between-grade’ options, but this was not validated. Furthermore, grading eccentric PVR jets can be difficult and prone to interobserver variability. In a study comparing PVR grading with magnetic resonance imaging and echocardiography, the rate of PVR was underestimated by echocardiography [24].

PVR is not frequent after SAVR. In the present study, the rate of moderate-to-severe PVR at discharge was 0.7%. However, significant PVR is rarely left untreated at the time of surgery. In fact, in case of significant PVR, surgeons are prone to perform a local surgical repair or replacement of the implanted prostheses during the index procedure. In a recent study, the rate of moderate-to-severe PVR after SAVR was 0.5% when endocarditis cases were excluded from the analysis [25]. In the intermediate-risk population of the PARTNER 2 trial, the rate of moderate-to-severe PVR after SAVR was 0.6% [9]. As in most similar studies, PVR results in SAVR patients are not highlighted and its prognostic impact not documented. Our study showed worse survival with mild and moderate-to-severe PVR after SAVR, but this difference was not significant in Cox proportional hazards analysis; still, the adjusted risk estimates suggest that mild and moderate-to-severe PVR had a tendency towards decreased survival also among SAVR patients.

Significance

This study adds data from unselected population with everyday medical practice. Compared to randomized controlled trials, the present study with its broader inclusion criteria, fewer exclusion criteria and a rather long follow-up might lead to greater generalizability of these findings in heterogeneous patient populations.

Limitations

The retrospective nature is the main limitation of this study. Secondly, echocardiographic exam was performed according to standard practice and its results might differ from studies with strict imaging protocols as in randomized controlled trials. Thirdly, we do not have data on late echocardiographic controls. Therefore, data on possible changes in PVR severity over time were not available for a throughout evaluation of this complication. Fourthly, we do not have data on mortality related to cardiovascular events. Finally, the incidence of moderate-to-severe PVR after SAVR was rather low and this may cause a type II error in our analyses. Indeed, the adjusted risk estimates suggest that patients with mild and moderate-to-severe PVR have a tendency towards decreased 4-year survival.

CONCLUSIONS

This study from an unselected nationwide series demonstrated that both mild and moderate-to-severe PVR after TAVR are associated with lower survival at mid-term. New-generation TAVR prostheses significantly reduced the risk of such a complication. PVR is not frequent after SAVR. However, a tendency towards lower survival was observed in patients with mild and moderate-to-severe PVR, also after SAVR.

SUPPLEMENTARY MATERIAL

Supplementary material is available at EJCTS online.

Funding

This study was supported by research grants to T.L. from the Finnish Foundation for Cardiovascular Research, the Finnish Medical Foundation and the Finnish Society of Angiology.

Conflict of interest: none declared.

Author contributions

Teemu Laakso: Conceptualization; Formal analysis; Writing—original draft. Mika Laine: Conceptualization; Writing—review & editing. Noriaki Moriyama: Conceptualization; Validation. Sebastian Dahlbacka: Investigation. Juhani Airaksinen: Investigation. Marko Virtanen: Investigation. Annastiina Husso: Investigation. Tuomas Tauriainen: Investigation. Matti Niemelä: Investigation. Timo Mäkikallio: Investigation. Antti Valtola: Investigation. Markku Eskola: Investigation. Tatu Juvonen: Investigation. Fausto Biancari: Data curation; Project administration; Writing—review & editing. Peter Raivio: Conceptualization; Writing—review & editing.

Reviewer information

European Journal of Cardio-Thoracic Surgery thanks Afksendiyos Kalangos, Tadashi Kitamura and the other, anonymous reviewer(s) for their contribution to the peer review process of this article.

REFERENCES

ABBREVIATIONS

 
• AS

   Aortic stenosis

 
• CI

   Confidence interval

 
• E-CABG

   European Coronary Artery Bypass Grafting

 
• EuroSCORE

   European System for Cardiac Operative Risk   Evaluation

 
• HR

   Hazard ratio

 
• OR

   Odds ratio

 
• PVR

   Paravalvular regurgitation

 
• SAVR

   Surgical aortic valve replacement

 
• TAVR

   Transcatheter aortic valve replacement